The present invention relates to a method for note taking information from documents. In particular, the invention relates to graphics containing machine readable spatial symbology.
Non-electronic methods for taking notes from books, documents, and other printed materials are generally slow or unreliable. Students, lawyers, and other people dependent upon prior authority have limited resources by which they may select desired portions of a document and reproduce them quickly and accurately. Generally, this requires marking selected passages from printed documents with tabs or highlighters and reproducing them manually at a later time. For example, U.S. Pat. No. 3,958,816 to Remmey, III discloses notation related book markers wherein double sided tabs may be adhered to pages of books. Numbers are written in opposite directions on opposite sides of the tab so that the tab may be affixed to a left or right page of a book to mark the page, and the point on the page, to which it refers. Key cards have numbers corresponding to the tabs. Notes are hand written on lines on the cards and the cards are attached to the front leaf of a book.
Recent developments in the area of optical character recognition (OCR) scanning, such as that disclosed in U.S. Pat. No. 5,920,877 to Kolster, make it possible to acquire discrete text strings and organize those strings in a preselected format for later use. Such devices, however, depend upon the scanner""s ability to recognize a character based on known OCR techniques, such as geometric OCR, which detects the printed character""s shape. Variations in fonts make it difficult to achieve total accuracy with geometric OCR.
U.S. Pat. No. 5,875,261 to Fitzpatrick et al. discloses a method of enhancing spatial character recognition by combining it with color coded OCR. Color coded OCR is a technique involving attempts to recognize a character based on color embedded in the character. For example, red may denote the letter xe2x80x9cAxe2x80x9d, whereas pink denotes the letter xe2x80x9caxe2x80x9d and yellow indicates the letter xe2x80x9cYxe2x80x9d. Color coding, by itself, may be subject to color processing errors involving color intensity, color density, color shifts, and color scanner misalignment. Anything effecting the spectrum, such as saturation or changes in hue, could also effect the processing. Consequently, Fitzpatrick et al. mixes color coding with geometric OCR and utilizes a hypothesis testing technique wherein the geometric OCR is used to generate a null hypothesis that a particular character has been recognized, and an alternative hypothesis that the character has not been recognized. Color coded OCR is then employed to select either the null hypothesis or the alternative hypothesis. This method increases accuracy of the scanning device, however, it is time consuming and requires color printing.
Barcode recognition and processing methods, such as those disclosed in U.S. Pat. No. 4,992,650 to Somerville, U.S. Pat. No. 5,227,616 to Lee, U.S. Pat. No. 5,229,584 to Erickson, and U.S. Pat. No. 5,451,760 to Renvall, provide optical scanning techniques with both speed and accuracy. Barcodes can be used for automatic reproduction of strings of printed marks and characters, however, barcode detracts from a printed page to be read by humans.